8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
1/40
The children
left behindA league table of inequality in child
well-being in the worlds rich countries
UNICEFInnocenti Research Centre
Report Card 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
2/40
Innocenti Report Card 9 was written by Peter Adamson.
Part 1 of the Report draws on the analysis carried out by Candace Currie,
Dorothy Currie, Leonardo Menchini, Dominic Richardson and
Chris Roberts, and presented in theInnocenti Working Paper2010-19
(available on the UNICEF Innocenti Research Centre (IRC) website:
www.unicef-irc.org).
ForReport Card 9 the Health Behaviour in School-aged Children (HBSC)
International Coordinating Centre provided the statistical results for the
analysis of inequality in childrens health. The Organisation for Economic
Co-operation and Development (OECD) provided distributional analysis of
child well-being indicators for material well-being and educational
outcomes. HBSC and the OECD are not responsible for the interpretation
of these results or other analysis in this report. The project was
coordinated by the UNICEF IRC.
The UNICEF Innocenti Research Centre would like to acknowledgethe generous support forReport Card 9 provided by the Andorran,
Australian, Belgian, German, Swiss and United Kingdom National
Committees for UNICEF.
Any part of thisInnocenti Report Card may be freely reproduced using
the following reference:
UNICEF (2010), The Children Left Behind: A league table of inequality
in child well-being in the worlds rich countries,Innocenti Report Card 9,
UNICEF Innocenti Research Centre, Florence.
TheReport Card series is designed to monitor and compare the
performance of economically advanced countries in securing the rights
of their children.
The UNICEF Innocenti Research Centre in Florence, Italy, was
established in 1988 to strengthen the research capability of the
United Nations Childrens Fund (UNICEF) and to support its advocacy
for children worldwide.
The Centre helps to identify and research current and future areas of
UNICEFs work. Its prime objectives are to improve international
understanding of issues relating to childrens rights and to help facilitate
the full implementation of the United Nations Convention on the Rights of
the Child in all countries.
The Centres publications are contributions to a global debate on child
rights issues and include a wide range of opinions. The views expressed
are those of the author and researchers and do not necessarily reect the
policies or views of UNICEF.
United Nations Childrens Fund (UNICEF), November 2010
UNICEF Innocenti Research Centre
Piazza SS. Annunziata, 12
50122 Florence, Italy
Tel: (+39) 055 2033 0
Fax: (+39) 055 2033 220
www.unicef-irc.org
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
3/40
UNICEF
Innocenti Research Centre
Whether in health, in education, or in material well-being, some
children will always fall behind the average. The critical question
is how far behind? Is there a point beyond which falling behind
is not inevitable but policy susceptible, not unavoidable butunacceptable, not inequality but inequity?
There are no widely agreed theoretical answers to these
questions. Report Card 9 seeks to stimulate debate on the issue
by introducing a common measure of bottom-end inequality.
This permits each countrys performance to be assessed according
to the standard of what the best-performing countries have been
able to achieve. Such a standard may not represent the best thatmay be aspired to in theory, but in practice it suggests a level
below which falling behind is manifestly not inevitable.
The Report Card series is premised on the belief that the true
measure of a nations standing is how well it attends to its children
their health and safety, their material security, their education
and socialization, and their sense of being loved, valued, and
included in the families and societies into which they are born.
Its common theme is that protecting children during their vital,vulnerable years of growth is both the mark of a civilized society
and the means of building a better future.
This ninth report in the series builds on previous issues by
focusing specically on those children in all OECD countries
who are at risk of being left behind of being neither included
nor protected by the wealthy societies in which they live.
I N N O C E N T I R E P O R T C A R D 9 1
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
4/40
Fig. 1a A league table of inequality in child well-being
The table summarizes the ndings of Report Card 9, ranking 24 OECD countries by their
performance in each of three dimensions of inequality in child well-being.
Fig. 1b The overall record
Figure 1b ranks each country by its overall
inequality record. Three points have been
awarded for a better than average
performance, 2 points for a performance
at or close to the OECD average, and1 point for a below average performance
(see note for denitions). Countries in
alphabetical order within groups.
Figs. 1a and 1b are limited to the 24 OECD countries with available data for all three
dimensions of inequality in child well-being.
Note: To compare the inequality performance of the featured countries in each
dimension of child well-being, inequality scores for the individual indicators usedare rst converted to standard scores (i.e. inequality is measured in standard
deviations from the OECD unweighted average). The standardized scores are then
averaged to arrive at an inequality score for each dimension. For purposes of Figs.
1a and 1b, inequality close to average is dened as a score within the range of
-0.5 to +0.5 standard deviations from the OECD average. inequality lower than
OECD average is dened as having a standard deviation score greater than +0.5
from the OECD unweighted average. inequality higher than OECD average is
dened as having a standard score of less than -0.5 from the OECD unweighted
average.
Source:See page 30 (Data forReport Card 9: the surveys) for data sources used in
the measurement of inequality in the different dimensions of childrens well-being.
A league table of inequality in child well-beingin the worlds rich countries
T C R N T N
Material well-being rank Education well-being rank Health well-being rank
Switzerland 1 Finland 1 Netherlands 1
Iceland 2 Ireland 2 Norway 2
Netherlands 3 Canada 3 Portugal 3
Denmark 4 Denmark 4 Germany 4
France 5 Poland 5 Switzerland 5
Finland 6 Hungary 6 Belgium 6Austria 7 Sweden 7 Ireland 7
Norway 8 Netherlands 8 Denmark 8
Sweden 9 Spain 9 Canada 9
Germany 10 Iceland 10 Czech Republic 10
Czech Republic 11 Norway 11 United Kingdom 11
Luxembourg 12 Switzerland 12 Slovakia 12
Ireland 13 United Kingdom 13 Austria 13
Spain 14 Portugal 14 Sweden 14
Belgium 15 Slovakia 15 France 15
Portugal 16 Luxembourg 16 Finland 16
Canada 17 Czech Republic 17 Iceland 17
Greece 18 Greece 18 Poland 18
United Kingdom 19 United States 19 Luxembourg 19
Italy 20 Germany 20 Greece 20
Poland 21 Italy 21 Spain 21
Hungary 22 Austria 22 United States 22
United States 23 France 23 Italy 23
Slovakia 24 Belgium 24 Hungary 24
inequality lower thanOC average
inequality close toOC average
inequality higher thanOC average
Higher score = greater equality
8Denmark
Finland
Netherlands
Switzerland
7
Iceland
IrelandNorway
Sweden
6
Austria
Canada
France
Germany
Poland
Portugal
5
Belgium
Czech Republic
Hungary
Luxembourg
Slovakia
Spain
United Kingdom
3Greece
Italy
United States
2 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
5/40
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
6/40
quality-of-life requires appropriate measures
of inequality with each of these measures
being signicant in itself and none claiming
absolute priority over the others, says the
Commission on the Measurement of
Economic Performance and Social
Progress established in 2008 by the
President of France.*
Figure 1a therefore compares 24
OECD countries according to their
performance in limiting bottom-endinequality in three dimensions of
childrens well-being. Its rankings
conrm the Commissions view that
no one indicator can stand as an
adequate proxy for the others.
Measuring the gap
Depending on the available data,
two different methods are proposed
for estimating how far behind
children are being allowed to fall.
The rst compares the position of
the child at the 10th percentile (i.e. the
child at a lower point than 90% of
children in the society) with the child
at the 50th percentile (the median
position). The degree of inequality is
measured by the gap between the two,
expressed as a percentage of the
median position.
The second method (employed when
survey data are not suitable for analysis
by percentiles) compares the level of
well-being of the child at the median
with the average level of all those who
fall below the median.
Different geographical and historical
circumstances may help to explain
different degrees of inequality. And it
is of course a truism that there will
always be a bottom 10% in any country
and that 50% of children will always fall
below the national median. In this
sense, a degree of falling behind is
obviously inevitable. The critical
question is how far behind? Is there
a point beyond which falling behind
is not inevitable but policy susceptible,
not unavoidable but unacceptable,
not inequality but inequity?
There are no widely agreed
theoretical answers to these questions.
But international comparison canhelp to establish practical answers by
measuring falling behind according
to the standard of what the best-
performing OECD countries have
already achieved. This benchmark may
not represent the very best that can be
aspired to, but it does establish a level
below which bottom-end inequality
is manifestly not inevitable.
If, for example, the gap in educational
achievement between students at
the 10th and 50th percentiles is
signicantly wider in France or
Belgium than in Finland or Ireland
(Figure 3d) then it seems clear that
the children at the 10th percentile in
French and Belgian schools are falling
further behind the median than is
necessary. The difference between the
best performing countries and the rest
of the OECD nations can therefore
be read as a minimum measure of the
extent to which falling behind is
policy-susceptible the extent to
which it is not unavoidable but unjust.
International comparison therefore sets
each nations performance not against
an abstract concept of equality but
against the practical benchmark of what
other nations at similar levels of
economic development have already
achieved. It therefore provides a realistic
measure of the scope for improvement.
* The Commission is chaired by Joseph Stiglitz, Amartya Sen and Jean-Paul Fitoussi.
4 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
7/40
Mexico
Chile (2006)
Republic of Korea (2007)
Australia
Other OECD countries
OECD average
Greece
Portugal
Spain
Canada (2005)
Italy
Poland
Belgium
United Kingdom
Slovakia
Germany
Ireland
Luxembourg
Hungary
Switzerland (2007)
Czech Republic
Finland
France (2007)
Netherlands
Sweden
Iceland
Austria
Denmark
Norway
61.5
60.8
59.4
51.0
46.9
56.6
56.2
56.0
55.7
54.1
51.2
50.6
50.1
48.9
48.1
47.4
46.4
44.6
44.3
43.7
41.9
41.6
41.5
41.2
40.2
40.0
39.7
39.4
0 10 20 30 40 50 60 70
Gap between the child at the 10th percentile and the child at the 50th percentile (as % of 50th percentile)
The rst of the three dimensions of
inequality in childrens well-being
considered here is inequality in
childrens material well-being.
Child poverty is about more than
poverty of income. It is also about
poverty of opportunity and expectation,
of cultural and educational resources,
of housing and neighbourhoods, of
parental care and time, of local services
and community resources. But fromthe childs point of view, these different
dimensions of poverty are rarely
separate. Family circumstance,
employment and income, health and
education systems, and the local
environment all play interacting roles
in determining well-being.
No internationally comparable data
are currently available to capture this
complexity. But rather than relying
on income data alone, inequality in
childrens material well-being is
measured here by three indicators
for which suitable data are available
household incomes, access to basic educational
resources, and housing living space.
Household income
Calculations of income inequality
for children are based on the disposable
incomes of households with children
aged 0 to 17 (after adding benets,
deducting taxes, and making an
adjustment for the economies of scale
available to larger families). To measure
inequality at the bottom-end of the
distribution, the income of the child
at the 50th percentile (the median)
is compared with the income of the
child at the 10th percentile (i.e. poorer
than 90% of children). How far behind
are the poorest children being allowed to
fall? is then measured by the gap
between the two.
As Figure 2a shows, household income
inequality for children is lowest in
Norway, with the Nordic countries and
the Netherlands taking six of the top
eight places in the table. At the other
extreme, Italy, Canada Spain, Portugal
and Greece are seen to have the highest
levels of child income inequality. Data
on household disposable income are not
available for the United States.*
Basic educational resources
The second measure used to compare
inequality in material well-being is
access to basic educational resources.
Again, the same question is asked
how far behind are the least advantaged
children being allowed to fall?
Fig. 2a nequality in material well-being: incomeThe chart shows inequality at the bottom-end of the distribution in disposable income
available to children in 27 OECD countries. Calculations are based on the incomes of
households with children aged 0 to 17 (after adding benets, deducting taxes, and making
an adjustment for the economies of scale available to larger families). For each country,
the measure of bottom-end inequality used is the gap between the income of the child at
the 50th percentile (the median level) and the income of the child at the 10th percentile
(i.e. the child who is poorer than 90% of children).
The bar chart shows how far the children at the 10th percentile are falling behind
(expressed as a percentage of median income in households with children).
Notes: Other OECD countries are listed separately because data limitations prevent their inclusion in the
overview tables for each dimension of child well-being. The OECD average is an unweighted average for the23 countries included in the main league table.
Sources: EU SILC 2008. Data for France are from EU SILC 2007. See page 30 (Data forReport Card 9: the surveys)
for more detailed notes on country data including sources for Australia, Canada, Chile, Mexico, the Republic of
Korea, and Switzerland.
M A T R A N A T
* Using gross (pre-tax) household income, the income available to the child at the 10th percentile in the
United States is approximately 70% below the income available to a child at the median.
I N N O C E N T I R E P O R T C A R D 9 5
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
8/40
Mexico
New Zealand
Turkey
Chile
Japan
Australia
Republic of Korea
OECD average
Slovakia
Greece
United Kingdom
Hungary
BelgiumUnited States
Germany
Poland
Norway
Ireland
Portugal
Spain
France
Iceland
Czech Republic
Italy
Sweden
Canada
Finland
AustriaLuxembourg
Netherlands
Switzerland
Denmark
22.5
22.1
22.0
21.9
19.9
18.9
14.2
15.2
25.9
22.6
21.0
20.7
19.919.2
18.5
18.1
16.3
16.2
16.1
14.9
14.3
14.0
13.9
13.3
12.0
11.3
11.3
10.99.5
8.8
8.1
7.9
Other OECD countries
0 105 15 20 25 30
Figure 2b attempts an answer by
drawing on survey data from the
Programme of International Student
Assessment (PISA).
In the 2006 PISA survey (see page 30),
a representative sample of 15-year-old
students in OECD countries was askedwhich of the following were available
in their own homes:
a desk
a quiet place to study
a computer for school work
educational software
an internet connection
a calculator
a dictionary
school textbooks.
The resulting scores registered on
a scale of 0 to 8 do not lend
themselves to analysis by percentile.
Inequality is therefore measured by the
gap between the score of the child at
the median and the average score of all
children who fall below the median.
The results are presented in Figure 2b.
The availability of computers and
internet access depends to some extent
on the level of economic development
in each country; even poor children
in very wealthy countries, for example,
may have access to most or all of the
items on the home educational
resources list. The median score
therefore differs from country to
country. But the focus here is on
inequality on the gap between the
median score (column 2) and the
average score below the median
(column 3). Column 4 shows the
difference between the two and the
chart represents the inequality gap as
a percentage of the median.
Northern European countries again
dominate the top of the table. The
lowest placed Nordic country, Norway,
posts an equality score close to the
average for the OECD as a whole.
At the foot of the table, the United
Kingdom, Greece, and Slovakia show
the highest levels of inequality in
access to basic educational resources.
Fig. 2b Inequality in material well-being: educational resources
15-year-olds students in each country were asked which of the following were available at home: a desk, a quiet place to study,
a computer for school work, educational software, an internet connection, a calculator, a dictionary, school textbooks.
Inequality was measured by comparing each countrys median score (column 2) with the average score of those below
the median (column 3). Column 4 shows the difference between the two as an absolute number of missing educational items.
The bar chart on the right shows the inequality gap (as a percentage of the median for each country).
Notes: Other OECD countries are listed separately because data limitations
prevent their inclusion in the overview tables for each dimension of child well-
being. The OECD average is an unweighted average for the 24 countries included
in the main league table.
Source: PISA 2006 (see page 30).
Educational items
(range 0-8)Median
Average
below the
median
Average
absolute
gap
Gap between the average below the median
and the median (as % of median)
Denmark 7 6.4 0.6
Switzerland 7 6.4 0.6
Netherlands 7 6.4 0.6
Luxembourg 7 6.3 0.7Austria 7 6.2 0.8
Finland 7 6.2 0.8
Canada 7 6.2 0.8
Sweden 7 6.2 0.8
Italy 7 6.1 0.9
Czech Republic 7 6.0 1.0
Iceland 8 6.9 1.1
France 7 6.0 1.0
Spain 7 6.0 1.0
Portugal 7 5.9 1.1
Ireland 7 5.9 1.1
Norway 8 6.7 1.3
Poland 7 5.7 1.3
Germany 8 6.5 1.5
United States 7 5.7 1.3Belgium 8 6.4 1.6
Hungary 7 5.6 1.4
United Kingdom 8 6.3 1.7
Greece 6 4.6 1.4
Slovakia 7 5.2 1.8
OECD average 7.2 6.1 1.1
Other OECD countries
Republic of Korea 7 6.0 1.0
Australia 8 6.5 1.5
Japan 6 4.8 1.2
Chile 5 3.9 1.1
Turkey 5 3.9 1.1
New Zealand 8 6.2 1.8
Mexico 5 3.9 1.1
6 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
9/40
Mexico
Chile (2006)
Australia
Other OECD countries
OECD average
Hungary
Italy
United States (2007)
Poland
Canada (2006)Luxembourg
Slovakia
United Kingdom
Sweden
Denmark
Austria
Czech Republic
Portugal
Finland
Belgium
Norway
Ireland
Netherlands
France (2007)
SpainGreece
Switzerland (2007)
Germany
Iceland
45.3
26.5
11.8
20.8
33.4
31.8
29.3
28.6
27.526.2
24.9
24.4
24.4
23.0
22.9
22.2
19.6
19.3
19.0
18.8
17.7
17.6
14.5
14.514.0
9.1
8.9
8.8
0 10 20 30 40 50
Living space
The third measure of material well-
being is living space dened as the
number of rooms per person in
households with children aged 0 to
17 (not counting corridors, kitchens,
and bathrooms). Although only an
approximate measure of housing
conditions, space in the home is a
constant and important factor in
young peoples lives. In the OECD
countries as a whole, one child in
three is estimated to be living in
overcrowded conditions.iii
Figure 2c draws again on survey data
to estimate inequality in living space.
As with educational resources, the
measure used is the gap between the
living space score at the median and the
average score of children below the
median. By this measure, Iceland,
Germany and Switzerland can be seen
to have the lowest levels of inequality in
childrens living space (along with
Australia, which is among the countries
excluded from the main tables because
data are not available for all three
dimensions of child well-being). At the
bottom of the table, inequality is highest
in the United States, Italy and Hungary.
Material inequality:
an overview
Figures 2d and 2e combine the three
measures used household income,
access to educational resources, and
living space. For each country, and
for each indicator, the inequality
scores have been set on a common
scale in which 100 represents the
OECD average and 10 represents
one standard deviation (a commonly
used measure of how spread out theitems being measured are in relation
to the average for the group as a
whole). The individual indicator
scores are then averaged to provide
the overview of inequality in
childrens material well-being
presented in Figure 2d.
Switzerland has the least inequality,
closely followed by Iceland and the
Netherlands.
Fig. 2c Inequality in material well-being: housing living space
Housing living space is dened as the number of rooms per person in households with children (not counting corridors, kitchens, and
bathrooms). Inequality is measured by the gap between the score at the median (column 2) and the average score of all children below the
median (column 3). Column 4 shows the difference between the two. The bar chart on the right shows the inequality gap (as a percentage
of the median).
Notes: Other OECD countries are listed separately because data limitations
prevent their inclusion in the overview tables for each dimension of child well-
being. The OECD average is an unweighted average for the 24 countries included
in the main league table.
Sources: EU SILC 2008. Data for France are from EU-SILC 2007. See page 30 (DataforReport Card 9: the surveys) for more detailed notes on individual country data
including sources for Australia, Canada, Chile, Mexico, the Republic of Korea,
Switzerland and the United States.
Living space
Rooms per personMedian
Average
below the
median
Average
absolute
gap
Gap between the average below the median
and the median (as % of median)
Iceland 1.00 0.91 0.09
Germany 1.00 0.91 0.09
Switzerland (2007) 1.00 0.91 0.09
Greece 0.80 0.69 0.11Spain 1.25 1.08 0.18
France (2007) 1.00 0.85 0.15
Netherlands 1.25 1.03 0.23
Ireland 1.25 1.03 0.23
Norway 1.20 0.97 0.23
Belgium 1.20 0.97 0.23
Finland 1.20 0.97 0.23
Portugal 1.00 0.80 0.20
Czech Republic 0.80 0.62 0.18
Austria 1.00 0.77 0.23
Denmark 1.20 0.92 0.28
Sweden 1.20 0.91 0.29
United Kingdom 1.20 0.91 0.29
Slovakia 0.75 0.56 0.19
Luxembourg 1.25 0.93 0.33Canada (2006) 1.50 1.10 0.41
Poland 0.67 0.47 0.19
United States (2007) 1.25 0.89 0.36
Italy 1.00 0.68 0.32
Hungary 0.75 0.50 0.25
OECD average 1.07 0.85 0.22
Other OECD countries
Australia 1.00 0.88 0.12
Chile (2006) 0.75 0.55 0.20
Mexico 0.50 0.28 0.23
I N N O C E N T I R E P O R T C A R D 9 7
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
10/40
Slovakia
United States
Hungary
Poland
Italy
United Kingdom
Greece
Canada
Portugal
Belgium
Spain
Ireland
Luxembourg
Czech Republic
Germany
Sweden
Norway
Austria
Finland
France
Denmark
Netherlands
Iceland
Switzerland
85 90 95 100 105 110 115
inequality higher than
OECD-24 average
inequality close to
OECD-24 average
inequality lower than
OECD-24 average
Slovakia
United States**
Hungary
Poland
Italy
United Kingdom
Greece
Canada
Portugal
Belgium
Spain
Ireland
Luxembourg
Czech Republic
Germany
Sweden
Norway
Austria
Finland
France
Denmark
Netherlands
Iceland
Switzerland
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0
inequality higher than
OECD-24 average
inequality lower than
OECD-24 average
A second summary table (Figure 2e)
shows the individual contributions of
the three indicators, allowing countries
to see their strengths and weaknesses.
Countries such as Germany, Belgium,
the United Kingdom, Greece and
Slovakia, for example, are let down by
higher than average inequality inaccess to basic educational resources.
Spain, Canada, Portugal and Greece
lose ranking places by virtue of higher
than average levels of household
income inequality.
These three measures of bottom-end
inequality in childrens material well-
being are neither ideal norcomprehensive. But they are the best
available for the purposes of
international comparison. Rather than
recording material well-being solely
by the percentage of children in
households below a given income
threshold, they attempt a more
rounded measure of how far behind
the least advantaged children are beingallowed to fall.
Fig. 2d nequality in material well-being: an overview
Figure 2d combines the three measures of inequality in childrens
material well-being (income, educational items, living space) into
an overview for the 24 OECD countries with available data. For
each country, the inequality scores of the three indicators of
material inequality have been standardized, combined and placed
on a common scale in which 100 represents the OECD unweighted
average and 10 is equal to one standard deviation.*
Fig. 2e nequality in material well-being: a breakdown
Figure 2e presents the same information as Fig 2d but shows the
individual contributions of the three inequality indicators used. For
each indicator, the length of the bar represents each countrys
distance above or below the OECD 24 average (again measured in
standard deviations above or below that average). This allows
countries to see individual strengths and weaknesses.
* A standard deviation is a measure of the spread of the distribution aroundits average.
Sources: See individual Figs. 2a, 2b, 2c. See also Figure 2e for the standardized
inequality measure used for the three individual indicators of inequality in
childrens material well-being.
** No data are available on household disposable income for the United States.Sources: See individual Figs. 2a, 2b, and 2c.
income
living space
educational items
8 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
11/40
Mexico
Chile
Japan
New Zealand
Turkey
Australia
Republic of Korea
Other OECD countries
OECD average
Belgium
Italy
Greece
Czech Republic
Germany
SlovakiaFrance
Austria
Norway
Luxembourg
Portugal
United Kingdom
Iceland
Poland
Hungary
Spain
Netherlands
Sweden
Switzerland
Canada
Ireland
Denmark
Finland
31.3
30.0
28.6
27.9
26.8
25.1
21.8
28.1
32.5
32.1
31.5
31.4
31.1
31.030.7
30.3
29.7
29.3
29.2
28.4
27.6
27.0
26.8
26.6
26.5
26.2
26.1
24.8
24.2
24.2
19.9
0 5 10 15 20 25 30 35
C A T O N A N A T
The second dimension of inequality
considered here is inequality in young
peoples educational achievements.
The data are drawn from the
Programme of International Student
Assessment (PISA) which regularly
tests a nationally representative sampleof 15-year-old students* in more than
40 countries. The aim is to test and
compare prociency in reading, maths
and science.
As with income, the inequality measure
used is the gap between test scores at
the 10th and 50th percentiles. Figures
3a, 3b and 3c present the results.
Figure 3d combines the three measures
into an overview. Again, each countrys
score on each indicator has been set
on a common scale in which 100
represents the unweighted OECD
average and 10 represents one standard
deviation above or below that average.
This allows each countrys
performance to be measured in
relation to both the average and thedegree of variability for the OECD
as a whole.
Fig. 3a nequality in reading literacy
Using PISA scores for reading literacy of 15-year-old students, Figure 3a measures educational inequality in each country by comparing
the score of the student at the 50th percentile (the median) with the score of the student at the 10th percentile (i.e. lower than 90% of all
scores). The bar chart shows the gap between the two (expressed as a percentage of the median).
Notes: Other OECD countries are listed separately because data limitationsprevent their inclusion in the overview tables for each dimension of child well-
being. The OECD average is an unweighted average for the 23 countries included
in the main league table. Reading literacy data for the USA are missing.
Source: PISA 2006 (see page 30).
Reading literacy
50th
percentile
(median) score
10th
percentile
score
Absolute gap
(50th percentile
10th percentile)
Gap between the child at the 10th percentile and the child
at the 50th percentile (as % of 50th percentile)
Finland 550 441 109
Denmark 499 378 121
Ireland 522 395 127
Canada 534 402 132
Switzerland 506 373 133
Sweden 513 378 135
Netherlands 515 379 136
Spain 468 343 125
Hungary 490 359 131
Poland 513 374 139
Iceland 491 356 135
United Kingdom 501 359 142
Portugal 479 339 140
Luxembourg 487 344 143
Norway 492 346 146
Austria 499 348 151
France 499 346 153Slovakia 473 326 147
Germany 508 350 158
Czech Republic 489 335 154
Greece 469 321 148
Italy 478 325 153
Belgium 515 347 168
OECD average 500 359 141
Other OECD countries
Republic of Korea 563 440 123
Australia 519 388 131
Turkey 450 330 120
New Zealand 528 381 146
Japan 505 361 144
Chile 443 310 133
Mexico 415 285 130
*The survey samples only 15-year-olds who are attending school. It may therefore not fully represent marginalized groups in some OECD countries.
I N N O C E N T I R E P O R T C A R D 9 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
12/40
Mexico
Chile
Turkey
New Zealand
Japan
Republic of Korea
Australia
Other OECD countries
OECD average
BelgiumAustria
Czech Republic
France
Greece
Germany
Portugal
Luxembourg
Slovakia
Switzerland
United States
Spain
Norway
Hungary
Sweden
United Kingdom
Iceland
Netherlands
Poland
Canada
DenmarkIreland
Finland
Italy
26.4
25.9
23.8
23.3
23.2
22.5
22.0
24.1
27.927.0
26.2
26.2
26.1
25.8
25.5
25.2
25.0
24.8
24.2
24.1
23.9
23.1
23.1
22.9
22.8
22.8
22.4
21.5
21.421.2
19.3
26.1
0 5 10 15 20 25 30
New Zealand
Japan
Australia
Chile
Mexico
Republic of Korea
Turkey
Other OECD countries
OECD average
United States
France
Belgium
United Kingdom
Luxembourg
Germany
Austria
Switzerland
Italy
Iceland
Greece
Netherlands
Norway
Czech Republic
Denmark
Portugal
Slovakia
Spain
Sweden
Ireland
Canada
Poland
Hungary
Finland
27.1
26.4
25.6
25.6
24.9
23.4
21.9
25.5
28.4
28.3
27.9
27.4
27.0
26.9
26.8
26.7
26.3
26.2
26.0
25.5
25.2
25.2
25.0
24.9
24.8
24.7
24.6
24.4
24.1
23.4
23.2
20.0
0 5 10 15 20 25 30
Fig. 3b nequality in maths literacy
Using PISA scores for maths literacy of 15-year-old students, Figure 3b measures educational inequality in each country by comparing
the score of the student at the 50th percentile (the median) with the score of the student at the 10th percentile (i.e. lower than 90% of all
scores). The bar chart shows the gap between the two (expressed as percentage of median).
Fig. 3c nequality in science literacy
Using PISA scores for science literacy of 15-year-old students, Figure 3c measures educational inequality in each country by comparing
the score of the student at the 50th percentile (the median) with the score of the student at the 10th percentile (i.e. lower than 90% of allscores). The bar chart shows the gap between the two as a percentage of the median.
Notes: Other OECD countries are listed separately because data limitationsprevent their inclusion in the overview tables for each dimension of child well-being.
The OECD average is an unweighted average for the 24 countries included in the
main league table. Source: PISA 2006 (see page 30).
Notes: Other OECD countries are listed separately because data limitationsprevent their inclusion in the overview tables for each dimension of child well-being.
The OECD average is an unweighted average for the 24 countries included in the
main league table. Source: PISA 2006 (see page 30).
Maths literacy
50th
percentile
(median) score
10th
percentile
score
Absolute gap
(50th percentile
10th percentile)
Gap between the child at the 10th percentile and the
child at the 50th percentile (as % of 50th percentile)
Finland 550 444 106
Ireland 503 396 107Denmark 514 404 110
Canada 529 416 113
Poland 495 384 111
Netherlands 534 412 121
Iceland 507 391 116
United Kingdom 494 381 113
Sweden 503 387 116
Hungary 490 377 113
Norway 490 373 117
Spain 482 366 116
United States 472 358 114
Switzerland 534 401 133
Slovakia 494 370 124
Luxembourg 492 368 124
Portugal 468 348 120
Germany 505 375 130
Greece 461 341 120
Italy 462 341 121
France 499 369 130
Czech Republic 510 376 134
Austria 511 373 138Belgium 528 381 147
OECD average 501 381 120
Other OECD countries
Australia 521 406 115
Republic of Korea 550 426 124
Japan 526 404 122
New Zealand 522 401 122
Turkey 415 316 99
Chile 408 302 106
Mexico 406 299 107
Science literacy
50th
percentile
(median) score
10th
percentile
score
Absolute gap
(50th percentile
10th percentile)
Gap between the child at the 10th percentile and the
child at the 50th percentile (as % of 50th percentile)
Finland 566 453 113
Hungary 506 388 117
Poland 498 381 117
Canada 540 410 130
Ireland 510 385 124
Sweden 505 381 124
Spain 491 370 121
Slovakia 489 368 121
Portugal 476 357 119
Denmark 498 373 125
Czech Republic 514 385 130
Norway 488 365 123
Netherlands 530 395 139
Greece 477 353 124
Iceland 493 364 129
Italy 477 351 126
Switzerland 516 378 138
Austria 516 378 138
Germany 521 381 140
Luxembourg 490 358 132
United Kingdom 518 376 142
Belgium 518 374 145
France 501 359 142
United States 488 349 139
OECD average 505 376 129
Other OECD countries
Turkey 416 325 91
Republic of Korea 526 403 123
Mexico 407 306 102
Chile 434 323 111
Australia 530 395 136
Japan 539 396 142
New Zealand 534 389 141
1 0 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
13/40
Belgium
France
Austria
Italy
Germany
United States
Greece
Czech Republic
Luxembourg
Slovakia
Portugal
United Kingdom
Switzerland
Norway
Iceland
Spain
Netherlands
Sweden
Hungary
Poland
Denmark
Canada
Ireland
Finland
80 85 90 95 100 105 110 115 120 125 130
inequality higher
than average
inequality close
to average
inequality lower
than average
Belgium
France
Austria
Italy
Germany
United States**
Greece
Czech Republic
Luxembourg
Slovakia
Portugal
United Kingdom
Switzerland
Norway
Iceland
Spain
Netherlands
Sweden
Hungary
Poland
Denmark
Canada
Ireland
Finland
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0
inequality higher than
OECD-24 average
inequality lower than
OECD-24 average
Figure 3e breaks down this overview
in order to show how bottom-end
inequality in the three different kinds
of literacy contributes to each
countrys overall inequality score.
No trade off
Such measurements serve to informtwo commonly contested issues.
First, they undermine the argument
that steady progress towards equality
of opportunity in education means
that differences in educational
outcomes are now mostly a reection
of the distribution of natural abilities.
As Figures. 3a, 3b and 3c show,
different OECD countries have very
different patterns of bottom-end
inequality in educational outcomes;and it is reasonable to assume that this
is the result not of differences in the
distribution of natural abilities but of
differences in policies which, over
time, limit the extent to which less
able students fall behind. Figure 3d, for
example, shows that lower-achieving
students in Finland, Ireland and
Canada are far less likely to fall a
long way behind their peers than are
students in Austria, France or Belgium.
The pattern of bottom-end inequality
in educational outcomes therefore
reects more than the lottery of birth
and circumstance. It may reect
Fig. 3d ducational inequality: an overview
Figure 3d combines the three measures of inequality in childrens
educational outcomes (in reading, maths and science literacy) into
an overview for 24 OECD countries. For each country, the scores
on the three indicators have been standardized, averaged, and
placed on a common scale in which 100 represents the OECD
unweighted average and 10 is equal to one standard deviation.*
Fig. 3e ducational inequality: a breakdown
Figure 3e presents the same information as Figure 3d but shows
the individual contributions of the three inequality indicators used.
For each indicator, the length of the bar represents each countrys
distance above or below the OECD 24 average (again measured in
standard deviations above or below that average). This allows
countries to see individual strengths and weaknesses.
* A standard deviation is a measure of the spread of the distribution around its
average.
Sources: See Figs. 3a, 3b, and 3c. See also Fig 3e for the standardized inequality
measure used for the three individual indicators of inequality in educational
well-being.
** Reading literacy data for the United States are missing.
Sources: See individual Figs. 3a, 3b, and 3c.
reading
sciences
maths
I N N O C E N T I R E P O R T C A R D 9 1 1
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
14/40
15 20 25 30 35
400
420
440
460
480
500
520
540
560
580
Medianscore
% gap between 50th and 10th percentiles
15 20 25 30 35
500
520
540
560
580
600
620
640
660
680
90th
percentiles
core
% gap between 50th and 10th percentiles
Notes: Blue vertical and horizontal lines indicate unweighted OECD average (30 countries). Trend line obtained by linear regression. Data for the United States are missing.
For country abbreviations see page 33.
Source: PISA 2006 (see page 30).
Notes: Blue vertical and horizontal lines indicate unweighted OECD average (30 countries). Trend line obtained by linear regression. Data for the United States are missing.For country abbreviations see page 33.
Source: PISA 2006 (see page 30).
differences in national efforts to
reduce socio-economic disadvantage.
Or it may reect efforts to weaken
the link between socio-economic
disadvantage and school
achievement (children whose
mothers did not complete secondary
school, for example, are atsubstantially greater r isk of having
low reading literacy scores, but that
risk is two or three times greater in
some countries than in others.)iv
It is likely, also, that different degrees
of inequality reect different
degrees of policy concern, over time,
for those at risk of falling behind.
Second, international comparisons of
inequality in educational outcomes also
inform the issue of whether a trade-off
must be made between investing in
low-achieving students and maximizing
the potential of those in the higher
reaches of the ability range. Figure 3f(i)
suggests an answer to this question byshowing that there is no relationship
between greater inequality and better
performance at the median. In fact the
most unequal countries tend towards
slightly lower scores at the 50th
percentile. The two countries with the
lowest bottom-end inequality in reading
literacy, Finland and South Korea, are
also the two countries with the highest
median levels of educational achievement.
A child born in either of these countries
therefore has both a lower chance of
falling a long way behind his or her peers
and a higher chance of scoring above the
average reading literacy mark for the
OECD as a whole.
Figure 3f(ii) shows that the point holds
when we look at performance of the
highest-achieving students. Again, the
countries with better results at the 90th
percentile of achievement tend to be
the countries with the lowest levels of
bottom-end inequality.
Fig. 3f(i) ottom-end inequality and median achievement
The chart compares inequality in reading literacy with median level scores for reading literacy in 30 OECD countries.
Fig. 3f(ii) ottom-end inequality and top-end achievement
The chart compares inequality in reading literacy with scores at the 90th percentile of achievement in 30 OECD countries.
1 2 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
15/40
Turkey
Other OECD countries
OECD average
United States
Greece
Luxembourg
Hungary
Italy
Poland
Iceland
Sweden
Spain
Finland
Canada
Slovakia
FranceCzech Republic
Belgium
Switzerland
United Kingdom
Norway
Ireland
Denmark
Germany
Portugal
Austria
Netherlands
27.8
23.6
28.0
27.0
26.9
26.8
26.3
25.8
25.8
25.5
24.3
24.3
24.3
23.6
23.323.1
23.1
22.9
22.4
21.5
20.8
20.7
20.5
20.4
20.4
19.2
0 5 10 15 20 25 30
A T N A T
The third and last dimension of child
well-being in which the data permit
cross national measurement of
inequality is health.
Again, three indicators are used:
childrens self-reported health
complaints; healthy eating; and
frequency of vigorous physical activity.
All three are well-established markers
for childrens current and future health.
The data are derived from the 2005-
2006 round ofHealth Behaviour in
School-aged Children, a World Health
Organization collaborative study
which regularly surveys the health
behaviours of schoolchildren at ages
11, 13 and 15 in 41 countries of
Europe and North America.
Self-reported health
Among many other questions,
participants in the HBSC survey were
asked how often in the previous six
months they had experienced the
following problems:
headache
stomach ache
feeling low
feeling irritable
feeling bad tempered
feeling nervous
having difculty getting to sleep
feeling dizzy.
The answers were transferred onto
a scale ranging from 0 (frequent
occurrences of all seven complaints)
to 28 (no health complaints).
Figure 4a uses these scores to estimate
the degree of inequality in childrens
self-reported health. As before, the
measure used is the gap between each
countrys median score (column 2) and
the average score of all children below
Fig. 4a ealth inequality: self-reported health complaints
The 2005-2006 HBSC survey (see page 30) asked 11, 13 and 15-year-old students how often in the previous six months they had
experienced the following problems: headache, stomach ache, feeling low, feeling irritable, feeling bad tempered, feeling nervous,
having difculty getting to sleep, feeling dizzy. The answers were transferred onto a scale ranging from 0 (frequent occurrences of
all seven complaints) to 28 (no health complaints).
Inequality was then measured by comparing each countrys median score (column 2) with the average score of those below the
median (column 3). Column 4 shows the difference between the two. The bar chart on the right shows the inequality gap (as a
percentage of the median).
Notes: Other OECD countries are listed separately because data limitationsprevent their inclusion in the overview tables for each dimension of child well-being.
The OECD average is an unweighted average for the 24 countries included in the
main league table. Source: HBSC 2005-2006 (see page 30).
Health complaints
(range 0-28)Median
Average
below the
median
Average
absolute
gap
Gap between the average below the median
and the median (as % of median)
Netherlands 25.0 20.2 4.8
Austria 25.0 19.9 5.1
Portugal 25.0 19.9 5.1
Germany 23.0 18.3 4.7
Denmark 24.0 19.0 5.0
Ireland 23.0 18.2 4.8
Norway 23.0 18.1 4.9
United Kingdom 22.0 17.1 4.9
Switzerland 22.0 17.0 5.0
Belgium 23.0 17.7 5.3
Czech Republic 21.0 16.2 4.8France 21.0 16.1 4.9
Slovakia 20.0 15.3 4.7
Canada 22.0 16.7 5.3
Finland 22.0 16.7 5.3
Spain 23.0 17.4 5.6
Sweden 22.0 16.4 5.6
Iceland 21.0 15.6 5.4
Poland 22.0 16.3 5.7
Italy 19.0 14.0 5.0
Hungary 21.0 15.4 5.6
Luxembourg 22.0 16.1 5.9
Greece 22.0 16.1 5.9
United States 22.0 15.8 6.2
OECD average 22.3 17.1 5.2
Other OECD countries
Turkey 18.0 13.0 5.0
I N N O C E N T I R E P O R T C A R D 9 1 3
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
16/40
Turkey
Other OECD countries
OECD average
Hungary
Finland
Iceland
Ireland
United States
Italy
Spain
Austria
Denmark
Germany
LuxembourgSlovakia
Greece
Switzerland
United Kingdom
Portugal
Norway
Czech Republic
Sweden
France
Poland
Canada
Belgium
Netherlands
38.1
42.6
50.5
49.2
47.1
46.7
46.5
45.9
45.8
45.1
44.5
44.5
44.243.5
42.5
41.9
41.8
40.8
40.3
39.7
39.0
37.8
37.1
36.6
35.4
35.4
0 10 20 30 40 50 60
the median (column 3). The inequality
gap is presented both as an absolute
difference between the two scores
(column 4) and as a bar chart showing
the bottom-end inequality gap as a
percentage of each countrys median.
Self-reporting has limitations as anindicator of health status. Cultural
differences, for example, may play a part
in explaining differences between each
countrys average score (although self-
reporting by children has been shown
to be a good predictor of adult health
outcomesv). But the focus here is not
on averages but on the inequalities
revealed by comparing each countrys
median score with the average score
below the median.
Again it is noticeable that the countries
with the highest median levels of health
the Netherlands, Austria, and Portugal
are also the countries with the
lowest levels of health inequality.
Healthy eating and
vigorous physical activity
The second and third indicators
available for the measurement of
bottom-end inequalities in childrenshealth are based on HBSC survey
data under the headings of healthy
eating and vigorous physical activity.
Healthy eating is basic to a childs
normal growth and development and
to long-term health. Unhealthy
eating, by contrast, is associated with
a wide range of immediate and
long-term health problems including
obesity, type 2 diabetes, and cardio-
vascular disease.vi A key componentof healthy eating is the inclusion of
fruit and vegetables in a childs
daily diet.
Regular exercise in adolescence
also brings short and long-term health
benets and is positively associated
with cognitive development,
emotional well-being, and even
academic achievement.vii For children
and adolescents, the World Health
Organization recommends60 minutes of moderate to vigorous
physical exercise every dayviii
(a recommendation that is not widely
followed; taking an unweighted
average of the 41 countries included
in the HBSC study, only 12%
of 15-year-old girls and 20% of
15-year-old boys report taking an
hour of moderate to vigorous
physical activity every dayix).
In both cases, HBSC survey datahave been translated into scores for
healthy eating (on a scale of 0 to 14)
and frequency of vigorous physical
Fig. 4b ealth inequality: healthy eating
The 2005-2006 HBSC survey asked 11, 13 and 15-year-old students how often they ate fruit and vegetables. The answers were converted
into a healthy eating score on a scale of 0 (no fruit or vegetable consumption) to 14 (daily consumption of both fruit and vegetables).
Inequality was then measured by comparing each countrys median score (column 2) with the average score of those below the
median (column 3). Column 4 shows the difference between the two. The bar chart on the right shows the inequality gap (as a
percentage of the median).
Notes: Other OECD countries are listed separately because data limitations
prevent their inclusion in the overview tables for each dimension of child well-being.The OECD average is an unweighted average for the 24 countries included in the
main league table. Source: HBSC 2005-2006 (see page 30).
Healthy eating
(range 0-14)Median
Average
below the
median
Average
absolute
gap
Gap between the average below the median
and the median (as % of median)
Netherlands 10.0 6.5 3.5
Belgium 10.0 6.5 3.5
Canada 10.0 6.3 3.7
Poland 8.5 5.3 3.2
France 8.5 5.3 3.2
Sweden 8.5 5.2 3.3
Czech Republic 8.5 5.1 3.4
Norway 8.5 5.1 3.4
Portugal 8.5 5.0 3.5
United Kingdom 10.0 5.8 4.2
Switzerland 10.0 5.8 4.2
Greece 8.5 4.9 3.6
Slovakia 8.5 4.8 3.7Luxembourg 8.5 4.7 3.8
Germany 8.5 4.7 3.8
Denmark 10.0 5.6 4.4
Austria 7.3 4.0 3.3
Spain 8.0 4.3 3.7
Italy 8.5 4.6 3.9
United States 8.5 4.5 4.0
Ireland 10.0 5.3 4.7
Iceland 8.5 4.5 4.0
Finland 8.0 4.1 3.9
Hungary 7.3 3.6 3.7
OECD average 8.8 5.1 3.7
Other OECD countries
Turkey 8.5 5.3 3.2
1 4 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
17/40
Turkey
Other OECD countries
OECD average
France
Spain
Italy
Poland
Hungary
Sweden
Belgium
United States
Austria
GreeceCanada
Czech Republic
United Kingdom
Luxembourg
Denmark
Slovakia
Iceland
Finland
Germany
Ireland
Norway
Switzerland
Netherlands
54.6
32.3
43.6
42.0
41.9
41.0
35.8
35.5
35.3
34.9
34.1
33.833.2
33.0
32.1
31.1
29.9
28.4
26.8
26.7
25.2
24.9
24.7
24.4
24.1
0 10 20 30 40 50 60
activity (on a scale of 0 to 11).
Following the pattern already
established, Figs. 4b and 4c measure
inequality by the gap between the
median score of each country and the
average score for all children below
the median. The bar charts again
show each countrys inequality gap asa percentage of the national median.
For healthy eating, the lowest level
of inequality is to be found in the
Netherlands, Belgium and Canada
and the highest in Iceland, Finland
and Hungary.
For vigorous physical activity, the
Netherlands again has least inequality,
closely followed by Switzerland and
Norway. The highest levels of bottom-end inequality are to be found in
France, Italy and Spain.
Health: an overview
Figure 4d combines the three measures
of bottom-end inequality in childrens
health onto a standardized common
scale. As the bar chart shows, the
Netherlands heads the table by a
distance (with the lowest inequality in
all three indicators). The United States,Italy and Hungary show the highest
levels of bottom-end inequality in
childrens health.
Figure 4e breaks down this overall
performance by showing the
contributions of the three individual
indicators. It allows countries like
France and Poland, for example, to see
that their position in the bottom half
of the table is brought about by high
levels of inequality in vigorous
physical activity. Ireland and Finland,
on the other hand, would both be
closer to the top of the table if it
were not for high levels of inequality
in healthy eating.
Statistics and children
This attempt at an international
comparison of inequality in different
dimensions of childrens well-beingis a work in progress. But its clear
overall message is that children are
falling signicantly further behind
in some countries than in others.
In particular, Denmark, Finland, the
Netherlands, and Switzerland are
leading the way in limiting how far
behind the least advantaged children
are allowed to fall.
Before discussing some of the
implications of these overallndings, two other concerns
should be acknowledged.
Fig. 4c ealth inequality: vigorous physical activity
The 2005-2006 HBSC survey asked 11, 13 and 15-year-old students about their exercise habits outside school hours, converting
the answers into a score for frequency of vigorous physical activity on a scale of 0 (no vigorous physical activity) to 11 (frequent
vigorous physical activity).
Inequality was then measured by comparing each countrys median score (column 2) with the average score of those below the
median (column 3). Column 4 shows the difference between the two. The bar chart on the right shows the inequality gap (as a
percentage of the median).
Notes: Data for Portugal are missing. Other OECD countries are listed separately
because data limitations prevent their inclusion in the overview tables for eachdimension of child well-being. The OECD average is an unweighted average for
the 23 countries included in the main league table. Source: HBSC 2005-2006 (see page 30).
Vigorous physical activity
(range 0-11)Median
Average
below the
median
Average
absolute
gap
Gap between the average below the median
and the median (as % of median)
Netherlands 8.0 6.1 1.9
Switzerland 7.0 5.3 1.7
Norway 7.0 5.3 1.7
Ireland 7.0 5.3 1.7
Germany 7.0 5.2 1.8
Finland 8.0 5.9 2.1
Iceland 7.0 5.1 1.9
Slovakia 8.0 5.7 2.3
Denmark 8.0 5.6 2.4
Luxembourg 7.0 4.8 2.2
United Kingdom 7.0 4.8 2.2
Czech Republic 6.0 4.0 2.0
Canada 8.0 5.3 2.7Greece 7.0 4.6 2.4
Austria 7.0 4.6 2.4
United States 7.0 4.6 2.4
Belgium 7.0 4.5 2.5
Sweden 7.0 4.5 2.5
Hungary 7.0 4.5 2.5
Poland 6.0 3.5 2.5
Italy 7.0 4.1 2.9
Spain 6.0 3.5 2.5
France 7.0 3.9 3.1
OECD average 7.1 4.8 2.3
Other OECD countries
Turkey 6.0 2.7 3.3
I N N O C E N T I R E P O R T C A R D 9 1 5
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
18/40
Hungary
Italy
United States
Spain
Greece
Luxembourg
Poland
Iceland
Finland
France
Sweden
Austria
Slovakia
United Kingdom
Czech Republic
Canada
Denmark
Ireland
Belgium
Switzerland
Germany
Portugal
Norway
Netherlands
85 90 95 100 105 110 115 120
inequality higher than
OECD-24 average
inequality close to
OECD-24 average
inequality lower than
OECD-24 average
Hungary
Italy
United States
Spain
Greece
Luxembourg
Poland
Iceland
Finland
France
Sweden
Austria
Slovakia
United Kingdom
Czech Republic
Canada
Denmark
Ireland
Belgium
Switzerland
Germany
Portugal**
Norway
Netherlands
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0
inequality higher than
OECD-24 average
inequality lower than
OECD-24 average
First, in measuring different
dimensions of childrens well-being it
is necessary to separate outcomes that
are rarely separated in childrens lives.
Multiple disadvantage is the norm
with each dimension intimately linked
and often mutually reinforcing at the
level of the individual childs life.
Second, the perennial danger of all
statistics is that in offering an overview
they can seem very distant from the
realities they seek to capture. And in
presenting these data, UNICEFs plea
is that the children themselves should
as far as possible be seen not as
statistics but as individual young
people, each with a name and a face,
each with needs and rights, each with
a personality and a potential, each
with a capacity to benet from and
contribute to the societies into which
they are born, and each with a keen
awareness of the norms of the societies
in which they live.
Fig. 4d ealth inequality: an overview
Figure 4d combines the three measures of inequality in childrens
health well-being (self-reported health complaints, healthy eating,
and vigorous physical activity) into an overview for the 24 OECD
countries with available data. For each country, the inequality
scores for the three indicators of health well-being have been
standardized, averaged and placed on a common scale in which
100 represents the OECD average and 10 is equal to one
standard deviation.*
Fig. 4e ealth inequality: a breakdown
Figure 4e presents the same information as Figure 4d but shows
the individual contributions of the three inequality indicators used.
For each indicator, the length of the bar represents each country's
distance above or below the OECD 24 average (again measured in
standard deviations above or below that average). This allows
countries to see individual strengths and weaknesses.
* A standard deviation is a measure of the spread of the distribution aroundits average.
Sources: See individual Figs. 4a, 4b, and 4c. See also Fig 4e for the standardized
inequality measure used for the three individual indicators of inequality in child
well-being.
** Data on vigorous physical activity for Portugal are missing.Sources: See individual Figs. 4a, 4b, and 4c.
vigorous physical activity
health complaints
healthy eating
1 6 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
19/40
A case has been made that national
averages are an insufcient guide to
national performance in meeting
childrens needs. Equity measures,and in particular measures of bottom
end inequality, are also needed. (The
UNICEF global report on Progress
for Children 2010makes a similar case
for the inclusion of equity measures
in monitoring the Millennium
Development Goals).
Secondly, it has been argued that
falling behind has many mutually
reinforcing dimensions and can not
be adequately represented by anysingle indicator. Policies to prevent
children from falling behind must
therefore address the different
dimensions of disadvantage
individually as well as collectively.
Such policies are largely a matter for
national research and debate. But an
international perspective can perhaps
offer some insights into this under-
researched area.
Equity in education
For the purposes of reducing bottom-
end inequality in childrens educational
achievement, for example, it is clear
that school admissions policies can
make a difference.
In all OECD countries where studies
have been conducted,x the average
socio-economic level of students in
a particular school has been found
to have an effect on educationalachievement that is over and above
the effects associated with the
socio-economic status of the individual
student. This nding strongly suggests
that pupils from lower socio-economic
backgrounds benet from attendingschools in which a wide range of
home backgrounds are represented.
Conversely, falling behind is
signicantly more likely when students
from homes of low socio-economic
status attend schools in which the
average socio-economic status is
also low.xi
The reasons for this school
composition effect are many. Schools
with low socio-economic proles maynd themselves struggling against lower
expectations on behalf of both staff and
students; the ethos and disciplinary
climate may be less conducive to
learning; pupil-teacher relations may
be less positive; parental involvement
and support may be weaker; and the
task of attracting and retaining the
most able teachers may be more
difcult. All of these are formidable
barriers to learning.
In many OECD countries there are
signicant numbers of schools in which
the average socio-economic prole is
below the 20th percentile of the socio-
economic distribution for the OECD
as a whole.xii In such cases, the school
composition effect is enlisted against
rather than in favour of those who are
already most at risk of educational
underachievement. The likely result is
an increase in bottom-end inequality.
Two obvious approaches may counter
this effect. First, the attempt can be
made to boost the performance of
low socio-economic status schools
(for example by increasing the
resources available to them andallowing them to offer extra incentives
to more able teachers). Second,
admission policies can be designed to
avoid the concentration of pupils from
disadvantaged backgrounds in low
socio-economic status schools. This
might be achieved, for example, by
admitting children in ability bands
without regard to socio-economic
background. Policies designed to
monitor and balance the socio-
economic prole of pupil intakemay also be important. As a 2006
report commissioned by UNESCO
has pointed out:
Countries with high levels of segregation
along socio-economic lines tend to have
lower overall performance and greater
disparities in performance between
students from high and low socio-economic
backgrounds
In countries with high levels of socio-economic segregation, policies that aimed
to reduce socio-economic segregation through
compensatory reforms would likely bring
considerable gains in raising and leveling
the learning bar.xiii
In practice, a combination of both
approaches will be necessary in
countries with high bottom end
inequality in educational outcomes
(shown in Figure 3d where the
bottom ve countries are Belgium,France, Austria, Italy and Germany).
Part 2
I N N O C E N T I R E P O R T C A R D 9 1 7
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
20/40
Resistance to such policies is common
and is often based, at least in part, on
fears that overall educational outcomes
might be adversely affected. But the
international comparisons set out in
Figs. 3f(i) and 3f(ii) suggest that lower
bottom-end inequality need not imply
any lowering of standards for high-achieving students. As the report for
UNESCO already cited concluded:
Successful schools tend to be those that
bolster the performance of those from less
advantaged backgrounds. Similarly, countries
that have the highest levels of performance
tend to be those that are successful in not
only raising the learning bar but also
leveling it.xiv
Equity in healthHigher than average inequality in
childrens health may also need to
be addressed by specic health sector
policies. Depending on context, such
policies might include: the targeting
of resources and outreach programmes
on those most at risk; the setting of
specic disparity reduction targets
for key health indicators such as
obesity, exercise, healthy eating and
infant and child mortality rates; and
increasing the reach and renement ofprevention policies designed to reduce
the impact of the health behaviours
that contribute most to bottom-end
inequality in health outcomes (such as
obesity, drug and alcohol abuse, and
smoking).
But there are also clear dangers in a
sectoral approach to reducing bottom-
end inequalities in childrens health.
It is tempting to target the bottom endof the distribution with policies aimed
at specic changes in lifestyles such as
promoting exercise and healthy eating
or reducing smoking or obesity levels.
But necessary as such programmes
are,* they cannot address the fact that
inequality in health outcomes, as in
educational outcomes, is principally
driven by socio-economic status.xv
A 2010 review of health inequalities in
the United Kingdom and what can
be done about them has this to say:
Inequalities in heath arise because of
inequalities in society in the conditions in
which people are born, grow, live, work, and
age. So close is the link between particular
social and economic features of society andthe distribution of health among the
population, that the magnitude of health
inequalities is a good marker of progress
towards creating a fairer society. Taking
action to reduce inequalities in health does
not require a separate health agenda, but
action across the whole of society.xvi
The signicance of the social gradient
in health has been demonstrated by
a steady ow of research ndings in
many OECD countries over recentyears.xvii Taking the three indicators of
inequality in childrens health used in
Part 1 of this report, for example, the
detailed HBSC data clearly show that
children of more afuent families take
more regular exercise, have healthier
eating habits, and report fewer health
problems.xviii Socio-economic status,
it is worth reminding ourselves, is
neither the choice nor the
responsibility of the child.
Among other studies, particularly
striking is the nding in Canada that
exposure to poverty in childhood
doubles the risk of death by age 55.xix
Similarly in the United States, socio-
economic status in childhood has been
shown to be a powerful predictor of
cardio-vascular disease in later life.xx
In Europe, the 2006 report Health
Inequalities: Europe in Proleconcludes
that, across the board, the poor have
shorter lives and more years of illhealth. Socio-economic inequalities in
health, says the reports author, Johan
Mackenbach of Rotterdam Universitys
Medical Centre, are unacceptable, and
represent one of Europes greatest challenges
for public health.xxi
Yet it is clear from the data presented
here and elsewhere that the relationship
between socio-economic status and
health is not xed. Being of low socio-
economic status clearly carries a greater
degree of risk in some countries than
in others. In most OECD countries,
for example, children born to parents
with low levels of education or into
homes with low socio-economic status
are more likely to die in the rst twelvemonths of life. Yet the steepness of this
social gradient in infant mortality
rates varies considerably from country
to county.xxii
Some countries, therefore, are clearly
doing a better job than others either
in reducing socio-economic inequalities
or in mitigating their impact on
childrens health and development. And
again it is the case that the countries
with the highest median levels of health the Netherlands, Austria, and Portugal
also have the lowest levels of health
inequality (Figure 4a). Conversely, the
countries whose children have the
lowest average levels of self-reported
health all tend to have higher-than-
average levels of health inequality.
The importance of income
Socio-economic status is therefore the
indispensable framework for policy
analysis of bottom-end inequality forchildren. For just as inequalities in
heath reect not only the effect of
health services but also the conditions
in which people are born, grow, live, work,
and age, so inequalities in educational
outcomes at age 15, for example, reveal
not only what happens in schools
but also the educational resources,
stimulation and encouragement that
surrounds a child from the earliest
weeks and months of life.
Policies designed to address specic
inequalities in health or education are
therefore likely to have limited impact
if they conne themselves to the health
and education sectors alone. The most
potent fact about children who fall
signicantly behind their peers is that,
by and large, they are the children of
families at the bottom end of the
socio-economic scale.
* In England, for example, smoking accounts for approximately half the difference in average life expectancy between the lowest and highest income groups.
(Michael Marmot (chair) 2010,Fair Societies, Healthy Lives, Strategic Review of Health Inequalities in England post 2010, p 10.)
1 8 I N N O C E N T I R E P O R T C A R D 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
21/40
Most OECD countries have adopted national poverty
lines based on a percentage of the nations median
income. The European Union, for example, draws the
poverty line at 60% of the median. The OECD uses
50% of the median.1
In some countries, the idea of relative poverty is still
a matter of dispute. Poverty, it is argued, should be
measured by absolute rather than relative standards.
In the United States, for example, the ofcial poverty
line is based on a multiple of the income required to
ensure an adequate diet.
But it could be argued that all denitions of poverty
other than the minimum required for sheer physicalsurvival are in fact relative denitions. Absolute
poverty in the not-too-distant past, for example, meant
that life was nasty, brutish and short. Absolute poverty
in the United States today means not being able to
afford a standard of living including standards of
nutrition, water supply, sanitation, health care and
transport far in advance of the standards enjoyed by
most of the worlds population for most of its history.
In this sense, even absolute denitions of poverty are
really relative denitions that eventually have to be
updated to take account of changing standards of what
is acceptable to the society as a whole. The questionthen becomes whether the denition should be updated
infrequently in anad hoc way or whether it should be
updated regularly and systematically for example by
tying it to the national median income.
In recent times, dening income poverty in relative
terms has become widely established, especially in the
European Union. In the United Kingdom, for example,
TheEconomistmagazine notes that A decade ago, the
prospect of the Conservatives accepting the idea of
relative poverty rather than an absolute measure of
want, such as a basket of goods that every household
should be able to afford would have been fanciful.
Nowadays, it is a reality.2
Box 1 Child poverty: a relative measure
1 In discussing child poverty rates, part 2 of this report follows
the method recommended by the OECD, drawing the poverty line
at 50% of national median household income. Household income
is taken to mean disposable household income, i.e. after taxes
and public transfers. This is then equivalized to take into account
the economies of scale available to different sizes of households
(using the square root of household size). The poverty line is
therefore dened as half of the median national disposable
equivalized income; the child poverty rate is then calculated on
the same basis but taking into account only households with
children aged 0-17.
2 Still with us, The Economist, 1 July 2010.
This is not to say that the idea is new. More than 200
years ago the founding father of modern economics
argued that poverty was a relative concept:
By necessaries I understand, not only the commodities
which are indispensably necessary for the support of
life, but whatever the custom of the country renders
it indecent for creditable people, even of the lowest
order, to be without. A linen shirt, for example, is,
strictly speaking, not a necessary of life. But in the
present times, through the greater part of Europe, a
creditable day-labourer would be ashamed to appear
in public without a linen shirt ... Custom, in the same
manner, has rendered leather shoes a necessary of life
in England. The poorest creditable person of either sex
would be ashamed to appear in public without them.
Under necessaries, therefore, I comprehend, not
only those things which nature, but those things which
the established rules of decency have rendered
necessary to the lowest rank of people.
Adam Smith, An Enquiry into the Nature and Causes
of the Wealth of Nations, Book 5, Chapter 2, 1776.
I N N O C E N T I R E P O R T C A R D 9 1 9
8/4/2019 Innocenti Report Card 9 - The Children Left Behind: A league table of inequality in child well-being in the world's ric
22/40
Action to prevent children from falling
behind in different dimensions of well-
being must therefore eventually come
face to face with the question of the
socio-economic gradient.
Income poverty
Socio-economic status is about morethan income. A familys past savings
and future prospects, housing and
neighbourhoods, levels of parental
education and expectation, and status
in relation to the mainstream or
established ethnic or linguistic
community all of these enter into
the socio-economic equation. Yet
of the available measures, the most
important single guide to, and
predictor of, a familys socio-economic
status remains its level of householdincome. Reducing bottom-end
inequality in incomes will not solve
all other problems, but it will make
their solution easier. Climbing the
socio-economic ladder is more feasible
if the rungs are closer together.
Reviewing many studies that show
a strong and consistent associationbetween relative income poverty and
falling behind, Susan Mayer makes
the point uninchingly:
Parental income is positively correlated
with virtually every dimension of child
well-being that social scientists measure,
and this is true in every country for which
we have data. The children of rich parents
are healthier, better behaved, happier and
better educated during their childhood and
wealthier when they have grown up thanare children from poor families.xxiii
Relative income poverty therefore
occupies a primus inter pares position
among the indicators of falling
behind. But monitoring of income
poverty that can exert such leverage
over the trajectories of childrens lives
is not simply a matter of calculating
what proportion of a nations childrenis growing up in households whose
income falls below a given threshold.
The depth, duration and timing of that
poverty in relation to the different
stages of a childs development may
also be critical. A 2007 Canadian
review of research into this issue
renes the point:
Studies that measure family income over
extended periods of time and include
changes in income and the depth of incomeinequality in their models and analysis
The time lag between the gathering of data through
sample surveys in different countries and the publishing
of that data in internationally comparable form is
approximately 3 years. Most of the data in this report
therefore apply to the years 2006 to 2008.
Normally, such a delay is no more than frustrating.
Socio-economic data of a kind used here tend to reect
long-term trends rather than year on year changes.
But much has changed in the world since 2008. Economic
recession has affected millions in the OECD countries.
The response of governments, whether by cuts in
spending or increases in taxation, is affecting many
millions more. Across the European Union, for example,unemployment is predicted to surpass the 10% mark by
the time this report is published. This means that
approximately 5 million more people will be unemployed
than before the crisis began. As joblessness is a principal
driver of poverty, it is likely that the material well-being of
children has deteriorated in some countries since 2008.
In the European Union, youth unemployment, in particular,
has risen from a pre-crisis level of under 15% to more
than 20% today.1 Many households have seen their
incomes fall and are facing difculties with debt
repayments. In the United States, as many as half of all
workers have taken a cut in pay or hours or suffered atleast temporary unemployment in the two half years since
Box 2 First call: children and recession
the crisis struck.2 Migrant workers and those on short-term
contracts are particularly vulnerable.
There may be worse to come. According to a report by
the European Union Social Protection Committee The full
impact of the crisis on labour markets and public nances
has yet to be faced.3
In other words, the snapshot of inequality in childrens
well-being presented in these pages is a snapshot taken
in good times.
No overall statistics are yet available to chart the impact
of recession on the children of the poorest families. But a
partial glimpse may be offered by the changing demands
on charities and government special assistance
programmes. The International F